Abstract
Asthma prevalence is increasing in adult and paediatric patients. In the present study, the association between different leisure time activities and new onset of wheezing was analysed in adolescents aged 16–18 yrs taking part in a questionnaire-based follow-up of the International Study on Asthma and Allergies in Childhood in Munich and Dresden, Germany.
Of the 3,785 adolescents who took part in the follow-up (76% response), 2,910 adolescents without earlier episodes of wheezing in childhood were included in the analyses. Of these, 330 (11.3%) reported new onset of wheeze during the previous 12 months.
In the bivariate analyses, exercising more than once per week or performing computer work >1 h·day−1 were inversely related to new onset of wheeze. In contrast, visiting discotheques on a regular basis increased the risk of new onset of wheeze (12.9 versus 9.9%). The observed inverse relationship between physical activity and new onset of wheeze was not an independent effect but mediated by differences in active smoking.
The association between physical activity and new onset of wheeze disappeared when active smoking was taken into account. However, the present data do not allow for determining whether smoking operated as a confounder or as an intermediate factor, i.e. whether physical activities prevented active smoking.
Asthma is a chronic inflammatory disease with increasing prevalence both in adult and paediatric patients 1. Several explanations for this phenomenon have been discussed, such as genetic factors, environmental influences or a western lifestyle 2. One aspect of western lifestyle is reduced physical activity. Several studies have shown an inverse association between the prevalence of asthma and physical activity.
Nystad et al. 3 investigated 2,188 schoolchildren aged 6–16 yrs and found that the prevalence of bronchial responsiveness increased with decreasing hours of exercise per week in asthmatic children. In a prospective community-based study with 757 asymptomatic children, Rasmussen et al. 4 reported an inverse relationship between physical fitness and the development of physician-diagnosed asthma. In a study involving twins, Huovinen et al. 5 showed that twins undergoing conditioning exercise had a lower risk of asthma than sedentary twins. Among males, leisure time physical activity seems to have a slightly protective effect on asthma risk 6. From a pathophysiological view, only limited data exist to explain this protective effect. Deep inspiration prevents bronchoconstriction in healthy subjects and induces bronchodilation in asthmatics 7. Several factors might influence the grade of physical activity, such as residential area (e.g. safety of the neighbourhood, rural versus urban area), distance to the next playground or park, or health belief of the parents 8. Due to the popularity and easy access to computer games, many children and adolescents tend to spend their leisure time at home.
The objective of the present study was to analyse the association of different types of leisure time activity with the development of wheeze in young adolescents. It was hypothesised that adolescents with less physical activity in their leisure time had a higher prevalence of wheezing compared with their active counterparts.
MATERIAL AND METHODS
Study population
The Study on Occupational Allergy Risks (SOLAR) I is a prospective population-based cohort study concerning the natural history of atopic and respiratory diseases performed in 2002 in Dresden and Munich, Germany. During 1995–1996, 6,399 subjects aged 9–11 yrs had originally participated in the International Study of Asthma and Allergies in Childhood (ISAAC) II 9. Of these, 4,893 subjects were located for the follow-up in 2002. A total of 3,785 (76%) returned the questionnaire and gave informed consent to combine the baseline and follow-up data. However, only 3,693 answered the question whether they had ever wheezed in their life. The present analysis is based on 2,910 adolescents who had never had an episode of wheezing in childhood when assessed at the first survey. At the time of follow-up, 330 (11.3%) of them reported wheezing within the previous 12 months (fig. 1⇓). Of these, 54 (16.3%) had doctor-diagnosed asthma.
Identification of the study population. Of the 6,262 children who participated in the study in 1995–1996, 3,785 took part in the follow-up in 2002. However, only 3,693 answered the question concerning current wheeze in the period 2002–2003. The present analysis was based on 2,910 adolescents who never had an episode of wheezing in childhood. #: 9–11-yr-olds; ¶: 15–17-yr-olds; +: included in the present analysis.
Questionnaire
All subjects had to complete a questionnaire 10 with 121 questions focusing on: genetic factors; respiratory symptoms and diseases; housing; pets; active and passive smoking; occupational choice and working environment; leisure time activity; physical development; and stress factors, as previously published 11. Subjects with asthma-related symptoms within the previous year were identified by the question: “Have you had wheezing within the previous 12 months?” Information on leisure time activity was categorised as follows: 1) four categories of sport frequency between not more than once per month and 3 times per week; 2) computer work and TV watching ≥1 h·day−1; and 3) visiting a discotheque (yes or no). The selection of the categories was performed with the objective of getting subgroups of approximately equal size. Potential confounders were: socioeconomic status (SES), as reflected by parental education; sex; body mass index (BMI; kg·m−2, >95th percentile); and passive and active smoking (defined as having smoked ≥1 yr in lifetime).
Statistical methods
First, the crude associations between the three types of variables (outcome: current wheeze; potential predictors: leisure time activities; and confounders) were calculated using 2xN tables and Chi-squared tests. The adjusted odds ratios (ORs) were then analysed for new onset of wheeze in multiple logistic regression models. One of the predictors was initially introduced and all five potential confounders were then included stepwise. Differences were considered significant at p-values <0.05.
RESULTS
When analysing the crude association of leisure time activity and new onset of wheeze, a significantly negative association was found between wheeze and increasing frequency of sport (p = 0.001 for linear trend) and computer work (p = 0.01; table 1⇓). Analysing the association of new onset of wheeze with different types of exercise (e.g. running, swimming, fitness centre) or with the number of hours spent doing sport, similar (not statistically significant) tendencies as for sport frequency were found (data not shown). In contrast, visiting a discotheque was significantly positively associated (p = 0.02) with current wheeze (table 1⇓). Subsequently, the association of different confounders with wheeze was investigated (table 1⇓). When analysing active smoking as a confounding factor, a highly significant positive association was found with wheeze, as well as with TV watching and visiting discotheques (p<0.001), while the association with frequency of sport and computer work was significantly negative (p<0.001). Passive smoking, female sex and low SES were further significantly positively associated with wheeze (p<0.001, p<0.001 and p = 0.08, respectively), while no statistical significance was found for TV watching or BMI (p = 0.71 and p = 0.77, respectively). In addition, more TV watching was associated with lower parental SES and personal education, while there was no clear association between computer work and social attributes (data not shown). In addition, low SES was significantly associated with smoking (p<0.0005). When stratifying the analysis for smokers and nonsmokers, no significant association was found between leisure time activity and wheeze (table 2⇓). Consequently, in the multivariate logistic regression analysis adjusted for smoking, the leisure time activities were not associated with a significant risk of new onset of wheeze. Passive and active smoking, however, remained highly significant risk factors with an adjusted OR of ∼2.5 for active smoking (table 3⇓). However, no effect modification was found by sex of the actively or passively smoking subject (data not shown). Besides smoking, only female sex implied a significant risk for new onset of wheeze.
Crude association of leisure time activity and different potential confounders with 12-months prevalence of wheeze
Leisure time activity and 12-months prevalence of wheeze
Adjusted odds ratios(OR) for incident wheeze
The ORs for leisure time activities remained significant in models with all other potential confounders (SES, sex, BMI or increase in BMI between the period 1995–1996 and 2002, early asthma, passive smoking) except active smoking (table 3⇑).
A parental history of asthma did not significantly increase the risk of new onset of wheezing.
DISCUSSION
In a recently published study 12, the association between smoking and the incidence of asthma during adolescence was investigated. The objective of the present study was to analyse the association between leisure time activities and onset of wheezing in young adolescents. To the best of the present authors’ knowledge, the current study is the first that simultaneously investigated the association of leisure time activity and active or passive smoking with wheezing in adolescents. The results showed a significant positive association between wheezing and visiting discotheques, and significant negative associations between wheezing and sport or computer work. However, active smoking was significantly associated with wheeze, as well as with all investigated leisure time activities, which means that active smoking is a substantial confounder of the association of leisure time activity and wheeze. This is an important aspect that has to be taken into account when analysing the effects of leisure time activities on asthma. Studies on adults reflecting the confounding role of smoking on leisure time activity and wheezing showed different results. Huovinen et al. 6 found no effect of smoking on the adult onset of asthma, while in another study 5, the same group reported an association only in females. In childhood and adolescence, both active and passive smoking are known risk factors for wheezing 13, 14. Therefore, a strong confounding effect appears obvious. The present data were evaluated cross-sectionally and do not necessarily reflect the chronological onset of smoking and wheezing in the participating subjects. For this reason, it is not possible to differentiate whether adolescents do not smoke because they are actively sporty or whether they are inactive due to active or passive smoking. In the former case, sport would be the true reason for the lower prevalence of wheezing.
Another important result of the present study was that adolescents working frequently with a computer smoked less; conversely, watching TV was associated positively with active smoking and negatively with social status and sport. To the present authors’ knowledge no data exists focusing on computer work and smoking in adolescents, and only limited data are available on smoking and television viewing. Results from two studies 15, 16 suggest that television viewing is a significant predictor of smoking volume, with a stronger relationship for higher levels of viewing. In addition, higher levels of television viewing appear to relate to earlier onset of smoking 17. So far, explanations for these phenomena can only be speculated. Although watching TV and working with the computer are sedentary activities, the former might be a more social activity where adolescents are joined by their smoking parents or friends 15. Results of a Finnish cohort study 18 and a study on 11–13-yr-old schoolchildren from Marseilles (France) 19 showed that adolescent smoking is associated with parental smoking. However, parental smoking habits were not registered in detail in the present study nor whether the adolescents had their own television set in their rooms. A further explanation might be that heavy smokers watch more television due to a poorer physical condition. However, this aspect might be more relevant in elderly subjects than in adolescents. Interestingly, lower social status was associated with higher frequency of TV watching. Although data focusing on the association between prevalence of smoking and SES are controversial, many studies show a strong inverse association between parental or own SES and smoking in children and adolescents 20, 21, especially in Germany 22–24. For further epidemiological studies on asthma and leisure time activity, the present results show that computer work cannot be used synonymously with TV watching in analysing the effects of leisure time activities on asthma.
Potentially due to the small number of participants with a parental history of asthma (125 with maternal and 112 with paternal diagnosis of asthma), this aspect did not significantly increase the risk of new onset of wheezing.
A limiting factor of the present study was that all data were collected by questionnaire, without physical examination of the subjects. However, the questionnaire used in the present prospective study was previously validated against clinical examinations 25. While the parents answered the baseline questions in 1995–1996, the adolescents themselves replied to the follow-up questionnaire. This process was carried out in order to increase the validity of the data 26.
In conclusion, the current study implicates that sport, computer work or TV watching are not independent risk factors for new onset of wheezing in adolescents. All risk factors are strongly associated with active smoking, which itself is a powerful predictor for the onset of wheezing. Thus, any study focusing on the association of leisure time activity and wheezing should include active smoking into the analysis. Computer work and TV watching seem to be associated with different socioeconomic status and lifestyle; therefore, they cannot be used synonymously with potential predictors for wheezing.
- Received November 24, 2006.
- Accepted June 14, 2007.
- © ERS Journals Ltd
References
